[FEA]: SoL on cub::Transform for custom/zip input and output iterator

### Is this a duplicate?

- [x] I confirmed there appear to be no [duplicate issues](https://github.com/NVIDIA/cccl/issues) for this request and that I agree to the [Code of Conduct](CODE_OF_CONDUCT.md)

### Area

CUB

### Is your feature request related to a problem? Please describe.

In cuOpt's PDLP Solver (linear programming solver) we are dependent on cub::Transform.

We call it at every single iteration of the solver here: https://github.com/NVIDIA/cuopt/blob/branch-25.10/cpp/src/linear_programming/pdhg.cu#L197

And you can find the functor here: https://github.com/NVIDIA/cuopt/blob/branch-25.10/cpp/src/linear_programming/utils.cuh#L102

I wanted to make sure with the CCCL team that this case of input/output iterator is properly handled by cub::Transform to reach SoL.

### Describe the solution you'd like

An SoL cub::Transform for such iterator (which might already be the case)

### Describe alternatives you've considered

_No response_

### Additional context

_No response_